Re: [syzbot] possible deadlock in io_sq_thread_finish

[Pavel Begunkov <asml.silence@gmail.com>]

On 07/03/2021 09:49, syzbot wrote:
> Hello,
> 
> syzbot found the following issue on:
> 
> HEAD commit:    a38fd874 Linux 5.12-rc2
> git tree:       upstream
> console output: https://syzkaller.appspot.com/x/log.txt?x=143ee02ad00000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=db9c6adb4986f2f2
> dashboard link: https://syzkaller.appspot.com/bug?extid=ac39856cb1b332dbbdda
> 
> Unfortunately, I don't have any reproducer for this issue yet.
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+ac39856cb1b332dbbdda@syzkaller.appspotmail.com

Legit error, park() might take an sqd lock, and then we take it again.
I'll patch it up

> 
> ============================================
> WARNING: possible recursive locking detected
> 5.12.0-rc2-syzkaller #0 Not tainted
> --------------------------------------------
> kworker/u4:7/7615 is trying to acquire lock:
> ffff888144a02870 (&sqd->lock){+.+.}-{3:3}, at: io_sq_thread_stop fs/io_uring.c:7099 [inline]
> ffff888144a02870 (&sqd->lock){+.+.}-{3:3}, at: io_put_sq_data fs/io_uring.c:7115 [inline]
> ffff888144a02870 (&sqd->lock){+.+.}-{3:3}, at: io_sq_thread_finish+0x408/0x650 fs/io_uring.c:7139
> 
> but task is already holding lock:
> ffff888144a02870 (&sqd->lock){+.+.}-{3:3}, at: io_sq_thread_park fs/io_uring.c:7088 [inline]
> ffff888144a02870 (&sqd->lock){+.+.}-{3:3}, at: io_sq_thread_park+0x63/0xc0 fs/io_uring.c:7082
> 
> other info that might help us debug this:
>  Possible unsafe locking scenario:
> 
>        CPU0
>        ----
>   lock(&sqd->lock);
>   lock(&sqd->lock);
> 
>  *** DEADLOCK ***
> 
>  May be due to missing lock nesting notation
> 
> 3 locks held by kworker/u4:7/7615:
>  #0: ffff888010469138 ((wq_completion)events_unbound){+.+.}-{0:0}, at: arch_atomic64_set arch/x86/include/asm/atomic64_64.h:34 [inline]
>  #0: ffff888010469138 ((wq_completion)events_unbound){+.+.}-{0:0}, at: atomic64_set include/asm-generic/atomic-instrumented.h:856 [inline]
>  #0: ffff888010469138 ((wq_completion)events_unbound){+.+.}-{0:0}, at: atomic_long_set include/asm-generic/atomic-long.h:41 [inline]
>  #0: ffff888010469138 ((wq_completion)events_unbound){+.+.}-{0:0}, at: set_work_data kernel/workqueue.c:616 [inline]
>  #0: ffff888010469138 ((wq_completion)events_unbound){+.+.}-{0:0}, at: set_work_pool_and_clear_pending kernel/workqueue.c:643 [inline]
>  #0: ffff888010469138 ((wq_completion)events_unbound){+.+.}-{0:0}, at: process_one_work+0x871/0x1600 kernel/workqueue.c:2246
>  #1: ffffc900023a7da8 ((work_completion)(&ctx->exit_work)){+.+.}-{0:0}, at: process_one_work+0x8a5/0x1600 kernel/workqueue.c:2250
>  #2: ffff888144a02870 (&sqd->lock){+.+.}-{3:3}, at: io_sq_thread_park fs/io_uring.c:7088 [inline]
>  #2: ffff888144a02870 (&sqd->lock){+.+.}-{3:3}, at: io_sq_thread_park+0x63/0xc0 fs/io_uring.c:7082
> 
> stack backtrace:
> CPU: 1 PID: 7615 Comm: kworker/u4:7 Not tainted 5.12.0-rc2-syzkaller #0
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> Workqueue: events_unbound io_ring_exit_work
> Call Trace:
>  __dump_stack lib/dump_stack.c:79 [inline]
>  dump_stack+0x141/0x1d7 lib/dump_stack.c:120
>  print_deadlock_bug kernel/locking/lockdep.c:2829 [inline]
>  check_deadlock kernel/locking/lockdep.c:2872 [inline]
>  validate_chain kernel/locking/lockdep.c:3661 [inline]
>  __lock_acquire.cold+0x14c/0x3b4 kernel/locking/lockdep.c:4900
>  lock_acquire kernel/locking/lockdep.c:5510 [inline]
>  lock_acquire+0x1ab/0x740 kernel/locking/lockdep.c:5475
>  __mutex_lock_common kernel/locking/mutex.c:946 [inline]
>  __mutex_lock+0x139/0x1120 kernel/locking/mutex.c:1093
>  io_sq_thread_stop fs/io_uring.c:7099 [inline]
>  io_put_sq_data fs/io_uring.c:7115 [inline]
>  io_sq_thread_finish+0x408/0x650 fs/io_uring.c:7139
>  io_ring_ctx_free fs/io_uring.c:8408 [inline]
>  io_ring_exit_work+0x82/0x9a0 fs/io_uring.c:8539
>  process_one_work+0x98d/0x1600 kernel/workqueue.c:2275
>  worker_thread+0x64c/0x1120 kernel/workqueue.c:2421
>  kthread+0x3b1/0x4a0 kernel/kthread.c:292
>  ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294
> 
> 
> ---
> This report is generated by a bot. It may contain errors.
> See https://goo.gl/tpsmEJ for more information about syzbot.
> syzbot engineers can be reached at syzkaller@googlegroups.com.
> 
> syzbot will keep track of this issue. See:
> https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
> 

-- 
Pavel Begunkov